Web surveys are everywhere, but their impact on the quality of survey data is unclear. A recent conference at Harvard University, Measuring from a Distance: The Emerging Science of Internet-Based Survey Research, brought top researchers in the field together to share insights and experimental evidence on the topic.
Frederick Conrad, director of the University of Michigan Program in Survey Methodology, based at the U-M Institute for Social Research, was one of the speakers.
Also speaking was Roger Tourangeau, co-author of The Science of Web Surveys with Conrad and ISR survey methodologist Mick Couper.
Conrad presented preliminary findings from a series of experiments funded by the National Science Foundation and the National Institutes of Health examining how the use of virtual interviewers and interactive prompts can affect responses and data quality.
In one experiment, Conrad presented data on the impact of speeding through questionnaires, and how textual prompts triggered by this undesirable behavior actually reduced its frequency.
“You seem to have responded very quickly,” the prompt begins. “Please be sure you have given the question sufficient thought to provide an accurate answer. Do you want to go back and reconsider your answer?”
This kind of prompt can be very effective at slowing down respondents, Conrad said. “But it may lead respondents to feel that they are being monitored so we need to see if it increases people’s tendency to provide socially desirable answers to sensitive questions.”
Conrad also presented data showing that prompts improved accuracy in respondents with intermediate levels of education – associate degrees or some college.
For more information about the Michigan Program in Survey Methodology, http://psm.isr.umich.edu/
For more information about the University of Michigan Institute for Social Research, http://home.isr.umich.edu